The objectives are to provide a better understanding of the anatomy, physiology and biochemistry of the gastric mucosal barrier and to elucidate those factors which disrupt and those which "tighten" the barrier. The permeability characteristics of normal and modified duodenal mucosa will be studied to ascertain whether a mucosal barrier similar to the gastric one exists and whether it can be altered by similar fat-soluble or surface-active agents. In the awake dog, gastric permeability will be studied in an animal with antrectomy and denervated fundic pouch using continuous perfusion of test solutions in a semi-open system. A similar canine preparation will be used to assess duodenal permeability with a duodenal pouch from the duct of Wirsung to the ligament of Trietz. On the basis of previous studies of validity, lithium will be used simultaneously with H ion as a marker for mucosal permeability. An ex vivo model of stripped rabbit gastric mucosa will also be used either in the short-circuited state or to assess changes in P.D. while bi-directional fluxes are measured in response to agents being tested for their effect on the barrier. Commonly used gastric irritants such as phenylbutazone and FeSO4 will be tested along with such experimentally ulcerogenic vasoactive materials as epinephrine and vasopressin. Possible barrier "tighteners" will be studied including Ca ions and carbenoxolone. The anatomical and biochemical aspects of the barrier will be investigated by electron microscopy and by chemical disruption of intracellular glycolysis and protein synthesis, respectively. Should a duodenal mucosal barrier similar to the gastric barrier be found, the effect of barrier disruption on normal duodenal function will be tested.